The Jerusalem Perinatal Study is a population-based research data bank based on mothers, fathers and offspring. In 1964-76 all Jewish births were surveyed, as were all births to Arab residents of West Jerusalem. Demographic data were recorded on the parents, infant deaths and congenital malformations, admissions to hospitals and obstetric complications. Approximately two thirds of the Jews were refugees, or their offspring, from Morocco, Algeria, Tunisia, Egypt, Iraq, Iran, Turkey, Yemen, Syria or Lebanon. Specific subsets of the mothers were interviewed in 1966-68 and 1975-76, adding information on consanguinity, antenatal health, body size, smoking, fertility, gynecologic variables and contraceptive use. After an average of 27 years (range 21-33) most offspring have completed their compulsory military service. This cohort may be one of the largest to include ethnic ancestry with obstetric and social data, which can link siblings and parents. We propose to link this database with Israel's Cancer Registry for follow-up studies of malignancies, as the parents and offspring age, and to prepare for the potential use of this cohort for cancer prevention trials. In this initial 4-year project we will undertake the checking and corrections needed for record linkage, describe losses to follow up, identify deaths and update the record linkage annually. Linkage in 1999 will identify malignancies diagnosed to the end of 1996, including an estimated 1846 cases in the 34,900 mothers and 485 in the 92,500 male and female offspring. We estimate there will be 440 cancers of breast and 122 of ovary, half of them to women of non-European origin. Focussing on breast cancer in mothers we will test for effects of gender of first offspring and explore ethnic differences in the effects of these and conventional risk factors. We will question whether changes in incidence of breast cancer and or other cancers can be predicted by births of offspring with major or minor birth defects, including neural tube defects and Down's syndrome, or by certain obstetric outcomes, including pre-eclampsia. The results may identify women who would benefit from intensive screening or provide new clues to the etiology of breast cancer.